Wall Street is turning Nvidia's AI chips into a new futures market: Chart of the Day
Nvidia (NVDA) forecasts 70% revenue growth in fiscal 2028, expanding beyond chip sales to include financing and rental revenue. CME Group plans to launch futures tied to the rental cost of Nvidia's H100 and B200 GPUs, settling against Silicon Data benchmarks, with H100 at $2.68/GPU-hour and B200 at $5.66/GPU-hour. Contracts do not reserve chips but pay based on benchmark price movements.
How this was made
The 30-second read
Why it matters
The futures contract aims to price compute power, turning Nvidia's chip rentals into a tradable asset class.
Market read
First disclosure of a dedicated futures market for AI compute, potentially reshaping pricing dynamics for Nvidia and the semiconductor sector.
What to watch
Potential volatility in rental rates and the need for reliable benchmark data could limit trader participation.
Background
Nvidia recently forecast 70% revenue growth for FY2028, highlighting the rapid expansion of AI spending.
Ticker impact
CME Group plans to launch futures on Nvidia H100 and B200 rental rates, a new tradable product tied to Nvidia's AI chips.
Potential short-term upside for NVDA as market participants price in increased liquidity and exposure to AI compute demand.
First report of CME futures linked to Nvidia chips; large‑cap issuer; novel market structure likely to attract speculative and hedging activity.
Market effects
May spur similar compute‑rental contracts for other AI chip makers, affecting the broader semiconductor sector.
U.S. futures markets gain a new product; could influence global AI compute pricing.
Creates a benchmark that could be referenced by AI‑heavy firms worldwide.
Counterpoint
If regulatory approval stalls, the futures may never launch, limiting any price impact on Nvidia.
Key entities
- CompanyNvidia
Designer of H100 and B200 GPUs, central to AI compute demand.
- ExchangeCME Group
Operator planning to list the AI compute futures.
- Data ProviderSilicon Data
Tracks hourly rental rates for Nvidia GPUs.





